#include <net/dsfield.h>
#include <net/inet_ecn.h>
#include <net/xfrm.h>
+#include <net/net_namespace.h>
+#include <net/netns/generic.h>
#ifdef CONFIG_IPV6
#include <net/ipv6.h>
static int ipgre_fb_tunnel_init(struct net_device *dev);
-static struct net_device *ipgre_fb_tunnel_dev;
+#define HASH_SIZE 16
+
+static int ipgre_net_id;
+struct ipgre_net {
+ struct ip_tunnel *tunnels[4][HASH_SIZE];
+
+ struct net_device *fb_tunnel_dev;
+};
/* Tunnel hash table */
will match fallback tunnel.
*/
-#define HASH_SIZE 16
#define HASH(addr) (((__force u32)addr^((__force u32)addr>>4))&0xF)
-static struct ip_tunnel *tunnels[4][HASH_SIZE];
-
-#define tunnels_r_l (tunnels[3])
-#define tunnels_r (tunnels[2])
-#define tunnels_l (tunnels[1])
-#define tunnels_wc (tunnels[0])
+#define tunnels_r_l tunnels[3]
+#define tunnels_r tunnels[2]
+#define tunnels_l tunnels[1]
+#define tunnels_wc tunnels[0]
static DEFINE_RWLOCK(ipgre_lock);
/* Given src, dst and key, find appropriate for input tunnel. */
-static struct ip_tunnel * ipgre_tunnel_lookup(__be32 remote, __be32 local, __be32 key)
+static struct ip_tunnel * ipgre_tunnel_lookup(struct net *net,
+ __be32 remote, __be32 local, __be32 key)
{
unsigned h0 = HASH(remote);
unsigned h1 = HASH(key);
struct ip_tunnel *t;
+ struct ipgre_net *ign = net_generic(net, ipgre_net_id);
- for (t = tunnels_r_l[h0^h1]; t; t = t->next) {
+ for (t = ign->tunnels_r_l[h0^h1]; t; t = t->next) {
if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr) {
if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
return t;
}
}
- for (t = tunnels_r[h0^h1]; t; t = t->next) {
+ for (t = ign->tunnels_r[h0^h1]; t; t = t->next) {
if (remote == t->parms.iph.daddr) {
if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
return t;
}
}
- for (t = tunnels_l[h1]; t; t = t->next) {
+ for (t = ign->tunnels_l[h1]; t; t = t->next) {
if (local == t->parms.iph.saddr ||
(local == t->parms.iph.daddr &&
ipv4_is_multicast(local))) {
return t;
}
}
- for (t = tunnels_wc[h1]; t; t = t->next) {
+ for (t = ign->tunnels_wc[h1]; t; t = t->next) {
if (t->parms.i_key == key && (t->dev->flags&IFF_UP))
return t;
}
- if (ipgre_fb_tunnel_dev->flags&IFF_UP)
- return netdev_priv(ipgre_fb_tunnel_dev);
+ if (ign->fb_tunnel_dev->flags&IFF_UP)
+ return netdev_priv(ign->fb_tunnel_dev);
return NULL;
}
-static struct ip_tunnel **__ipgre_bucket(struct ip_tunnel_parm *parms)
+static struct ip_tunnel **__ipgre_bucket(struct ipgre_net *ign,
+ struct ip_tunnel_parm *parms)
{
__be32 remote = parms->iph.daddr;
__be32 local = parms->iph.saddr;
h ^= HASH(remote);
}
- return &tunnels[prio][h];
+ return &ign->tunnels[prio][h];
}
-static inline struct ip_tunnel **ipgre_bucket(struct ip_tunnel *t)
+static inline struct ip_tunnel **ipgre_bucket(struct ipgre_net *ign,
+ struct ip_tunnel *t)
{
- return __ipgre_bucket(&t->parms);
+ return __ipgre_bucket(ign, &t->parms);
}
-static void ipgre_tunnel_link(struct ip_tunnel *t)
+static void ipgre_tunnel_link(struct ipgre_net *ign, struct ip_tunnel *t)
{
- struct ip_tunnel **tp = ipgre_bucket(t);
+ struct ip_tunnel **tp = ipgre_bucket(ign, t);
t->next = *tp;
write_lock_bh(&ipgre_lock);
write_unlock_bh(&ipgre_lock);
}
-static void ipgre_tunnel_unlink(struct ip_tunnel *t)
+static void ipgre_tunnel_unlink(struct ipgre_net *ign, struct ip_tunnel *t)
{
struct ip_tunnel **tp;
- for (tp = ipgre_bucket(t); *tp; tp = &(*tp)->next) {
+ for (tp = ipgre_bucket(ign, t); *tp; tp = &(*tp)->next) {
if (t == *tp) {
write_lock_bh(&ipgre_lock);
*tp = t->next;
}
}
-static struct ip_tunnel * ipgre_tunnel_locate(struct ip_tunnel_parm *parms, int create)
+static struct ip_tunnel * ipgre_tunnel_locate(struct net *net,
+ struct ip_tunnel_parm *parms, int create)
{
__be32 remote = parms->iph.daddr;
__be32 local = parms->iph.saddr;
struct ip_tunnel *t, **tp, *nt;
struct net_device *dev;
char name[IFNAMSIZ];
+ struct ipgre_net *ign = net_generic(net, ipgre_net_id);
- for (tp = __ipgre_bucket(parms); (t = *tp) != NULL; tp = &t->next) {
+ for (tp = __ipgre_bucket(ign, parms); (t = *tp) != NULL; tp = &t->next) {
if (local == t->parms.iph.saddr && remote == t->parms.iph.daddr) {
if (key == t->parms.i_key)
return t;
if (!dev)
return NULL;
+ dev_net_set(dev, net);
+
if (strchr(name, '%')) {
if (dev_alloc_name(dev, name) < 0)
goto failed_free;
goto failed_free;
dev_hold(dev);
- ipgre_tunnel_link(nt);
+ ipgre_tunnel_link(ign, nt);
return nt;
failed_free:
static void ipgre_tunnel_uninit(struct net_device *dev)
{
- ipgre_tunnel_unlink(netdev_priv(dev));
+ struct net *net = dev_net(dev);
+ struct ipgre_net *ign = net_generic(net, ipgre_net_id);
+
+ ipgre_tunnel_unlink(ign, netdev_priv(dev));
dev_put(dev);
}
static void ipgre_err(struct sk_buff *skb, u32 info)
{
-#ifndef I_WISH_WORLD_WERE_PERFECT
-/* It is not :-( All the routers (except for Linux) return only
+/* All the routers (except for Linux) return only
8 bytes of packet payload. It means, that precise relaying of
ICMP in the real Internet is absolutely infeasible.
}
read_lock(&ipgre_lock);
- t = ipgre_tunnel_lookup(iph->daddr, iph->saddr, (flags&GRE_KEY) ? *(((__be32*)p) + (grehlen>>2) - 1) : 0);
+ t = ipgre_tunnel_lookup(dev_net(skb->dev), iph->daddr, iph->saddr,
+ (flags&GRE_KEY) ?
+ *(((__be32*)p) + (grehlen>>2) - 1) : 0);
if (t == NULL || t->parms.iph.daddr == 0 ||
ipv4_is_multicast(t->parms.iph.daddr))
goto out;
out:
read_unlock(&ipgre_lock);
return;
-#else
- struct iphdr *iph = (struct iphdr*)dp;
- struct iphdr *eiph;
- __be16 *p = (__be16*)(dp+(iph->ihl<<2));
- const int type = icmp_hdr(skb)->type;
- const int code = icmp_hdr(skb)->code;
- int rel_type = 0;
- int rel_code = 0;
- __be32 rel_info = 0;
- __u32 n = 0;
- __be16 flags;
- int grehlen = (iph->ihl<<2) + 4;
- struct sk_buff *skb2;
- struct flowi fl;
- struct rtable *rt;
-
- if (p[1] != htons(ETH_P_IP))
- return;
-
- flags = p[0];
- if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) {
- if (flags&(GRE_VERSION|GRE_ROUTING))
- return;
- if (flags&GRE_CSUM)
- grehlen += 4;
- if (flags&GRE_KEY)
- grehlen += 4;
- if (flags&GRE_SEQ)
- grehlen += 4;
- }
- if (len < grehlen + sizeof(struct iphdr))
- return;
- eiph = (struct iphdr*)(dp + grehlen);
-
- switch (type) {
- default:
- return;
- case ICMP_PARAMETERPROB:
- n = ntohl(icmp_hdr(skb)->un.gateway) >> 24;
- if (n < (iph->ihl<<2))
- return;
-
- /* So... This guy found something strange INSIDE encapsulated
- packet. Well, he is fool, but what can we do ?
- */
- rel_type = ICMP_PARAMETERPROB;
- n -= grehlen;
- rel_info = htonl(n << 24);
- break;
-
- case ICMP_DEST_UNREACH:
- switch (code) {
- case ICMP_SR_FAILED:
- case ICMP_PORT_UNREACH:
- /* Impossible event. */
- return;
- case ICMP_FRAG_NEEDED:
- /* And it is the only really necessary thing :-) */
- n = ntohs(icmp_hdr(skb)->un.frag.mtu);
- if (n < grehlen+68)
- return;
- n -= grehlen;
- /* BSD 4.2 MORE DOES NOT EXIST IN NATURE. */
- if (n > ntohs(eiph->tot_len))
- return;
- rel_info = htonl(n);
- break;
- default:
- /* All others are translated to HOST_UNREACH.
- rfc2003 contains "deep thoughts" about NET_UNREACH,
- I believe, it is just ether pollution. --ANK
- */
- rel_type = ICMP_DEST_UNREACH;
- rel_code = ICMP_HOST_UNREACH;
- break;
- }
- break;
- case ICMP_TIME_EXCEEDED:
- if (code != ICMP_EXC_TTL)
- return;
- break;
- }
-
- /* Prepare fake skb to feed it to icmp_send */
- skb2 = skb_clone(skb, GFP_ATOMIC);
- if (skb2 == NULL)
- return;
- dst_release(skb2->dst);
- skb2->dst = NULL;
- skb_pull(skb2, skb->data - (u8*)eiph);
- skb_reset_network_header(skb2);
-
- /* Try to guess incoming interface */
- memset(&fl, 0, sizeof(fl));
- fl.fl4_dst = eiph->saddr;
- fl.fl4_tos = RT_TOS(eiph->tos);
- fl.proto = IPPROTO_GRE;
- if (ip_route_output_key(&init_net, &rt, &fl)) {
- kfree_skb(skb2);
- return;
- }
- skb2->dev = rt->u.dst.dev;
-
- /* route "incoming" packet */
- if (rt->rt_flags&RTCF_LOCAL) {
- ip_rt_put(rt);
- rt = NULL;
- fl.fl4_dst = eiph->daddr;
- fl.fl4_src = eiph->saddr;
- fl.fl4_tos = eiph->tos;
- if (ip_route_output_key(&init_net, &rt, &fl) ||
- rt->u.dst.dev->type != ARPHRD_IPGRE) {
- ip_rt_put(rt);
- kfree_skb(skb2);
- return;
- }
- } else {
- ip_rt_put(rt);
- if (ip_route_input(skb2, eiph->daddr, eiph->saddr, eiph->tos, skb2->dev) ||
- skb2->dst->dev->type != ARPHRD_IPGRE) {
- kfree_skb(skb2);
- return;
- }
- }
-
- /* change mtu on this route */
- if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED) {
- if (n > dst_mtu(skb2->dst)) {
- kfree_skb(skb2);
- return;
- }
- skb2->dst->ops->update_pmtu(skb2->dst, n);
- } else if (type == ICMP_TIME_EXCEEDED) {
- struct ip_tunnel *t = netdev_priv(skb2->dev);
- if (t->parms.iph.ttl) {
- rel_type = ICMP_DEST_UNREACH;
- rel_code = ICMP_HOST_UNREACH;
- }
- }
-
- icmp_send(skb2, rel_type, rel_code, rel_info);
- kfree_skb(skb2);
-#endif
}
static inline void ipgre_ecn_decapsulate(struct iphdr *iph, struct sk_buff *skb)
}
read_lock(&ipgre_lock);
- if ((tunnel = ipgre_tunnel_lookup(iph->saddr, iph->daddr, key)) != NULL) {
+ if ((tunnel = ipgre_tunnel_lookup(dev_net(skb->dev),
+ iph->saddr, iph->daddr, key)) != NULL) {
secpath_reset(skb);
skb->protocol = *(__be16*)(h + 2);
.saddr = tiph->saddr,
.tos = RT_TOS(tos) } },
.proto = IPPROTO_GRE };
- if (ip_route_output_key(&init_net, &rt, &fl)) {
+ if (ip_route_output_key(dev_net(dev), &rt, &fl)) {
tunnel->stat.tx_carrier_errors++;
goto tx_error;
}
.tos = RT_TOS(iph->tos) } },
.proto = IPPROTO_GRE };
struct rtable *rt;
- if (!ip_route_output_key(&init_net, &rt, &fl)) {
+ if (!ip_route_output_key(dev_net(dev), &rt, &fl)) {
tdev = rt->u.dst.dev;
ip_rt_put(rt);
}
}
if (!tdev && tunnel->parms.link)
- tdev = __dev_get_by_index(&init_net, tunnel->parms.link);
+ tdev = __dev_get_by_index(dev_net(dev), tunnel->parms.link);
if (tdev) {
hlen = tdev->hard_header_len;
int err = 0;
struct ip_tunnel_parm p;
struct ip_tunnel *t;
+ struct net *net = dev_net(dev);
+ struct ipgre_net *ign = net_generic(net, ipgre_net_id);
switch (cmd) {
case SIOCGETTUNNEL:
t = NULL;
- if (dev == ipgre_fb_tunnel_dev) {
+ if (dev == ign->fb_tunnel_dev) {
if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p))) {
err = -EFAULT;
break;
}
- t = ipgre_tunnel_locate(&p, 0);
+ t = ipgre_tunnel_locate(net, &p, 0);
}
if (t == NULL)
t = netdev_priv(dev);
if (!(p.o_flags&GRE_KEY))
p.o_key = 0;
- t = ipgre_tunnel_locate(&p, cmd == SIOCADDTUNNEL);
+ t = ipgre_tunnel_locate(net, &p, cmd == SIOCADDTUNNEL);
- if (dev != ipgre_fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
+ if (dev != ign->fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
if (t != NULL) {
if (t->dev != dev) {
err = -EEXIST;
err = -EINVAL;
break;
}
- ipgre_tunnel_unlink(t);
+ ipgre_tunnel_unlink(ign, t);
t->parms.iph.saddr = p.iph.saddr;
t->parms.iph.daddr = p.iph.daddr;
t->parms.i_key = p.i_key;
t->parms.o_key = p.o_key;
memcpy(dev->dev_addr, &p.iph.saddr, 4);
memcpy(dev->broadcast, &p.iph.daddr, 4);
- ipgre_tunnel_link(t);
+ ipgre_tunnel_link(ign, t);
netdev_state_change(dev);
}
}
if (!capable(CAP_NET_ADMIN))
goto done;
- if (dev == ipgre_fb_tunnel_dev) {
+ if (dev == ign->fb_tunnel_dev) {
err = -EFAULT;
if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
goto done;
err = -ENOENT;
- if ((t = ipgre_tunnel_locate(&p, 0)) == NULL)
+ if ((t = ipgre_tunnel_locate(net, &p, 0)) == NULL)
goto done;
err = -EPERM;
- if (t == netdev_priv(ipgre_fb_tunnel_dev))
+ if (t == netdev_priv(ign->fb_tunnel_dev))
goto done;
dev = t->dev;
}
.tos = RT_TOS(t->parms.iph.tos) } },
.proto = IPPROTO_GRE };
struct rtable *rt;
- if (ip_route_output_key(&init_net, &rt, &fl))
+ if (ip_route_output_key(dev_net(dev), &rt, &fl))
return -EADDRNOTAVAIL;
dev = rt->u.dst.dev;
ip_rt_put(rt);
struct ip_tunnel *t = netdev_priv(dev);
if (ipv4_is_multicast(t->parms.iph.daddr) && t->mlink) {
struct in_device *in_dev;
- in_dev = inetdev_by_index(dev->nd_net, t->mlink);
+ in_dev = inetdev_by_index(dev_net(dev), t->mlink);
if (in_dev) {
ip_mc_dec_group(in_dev, t->parms.iph.daddr);
in_dev_put(in_dev);
dev->flags = IFF_NOARP;
dev->iflink = 0;
dev->addr_len = 4;
+ dev->features |= NETIF_F_NETNS_LOCAL;
}
static int ipgre_tunnel_init(struct net_device *dev)
return 0;
}
-static int __init ipgre_fb_tunnel_init(struct net_device *dev)
+static int ipgre_fb_tunnel_init(struct net_device *dev)
{
struct ip_tunnel *tunnel = netdev_priv(dev);
struct iphdr *iph = &tunnel->parms.iph;
+ struct ipgre_net *ign = net_generic(dev_net(dev), ipgre_net_id);
tunnel->dev = dev;
strcpy(tunnel->parms.name, dev->name);
tunnel->hlen = sizeof(struct iphdr) + 4;
dev_hold(dev);
- tunnels_wc[0] = tunnel;
+ ign->tunnels_wc[0] = tunnel;
return 0;
}
static struct net_protocol ipgre_protocol = {
.handler = ipgre_rcv,
.err_handler = ipgre_err,
+ .netns_ok = 1,
};
+static void ipgre_destroy_tunnels(struct ipgre_net *ign)
+{
+ int prio;
-/*
- * And now the modules code and kernel interface.
- */
+ for (prio = 0; prio < 4; prio++) {
+ int h;
+ for (h = 0; h < HASH_SIZE; h++) {
+ struct ip_tunnel *t;
+ while ((t = ign->tunnels[prio][h]) != NULL)
+ unregister_netdevice(t->dev);
+ }
+ }
+}
-static int __init ipgre_init(void)
+static int ipgre_init_net(struct net *net)
{
int err;
+ struct ipgre_net *ign;
- printk(KERN_INFO "GRE over IPv4 tunneling driver\n");
+ err = -ENOMEM;
+ ign = kzalloc(sizeof(struct ipgre_net), GFP_KERNEL);
+ if (ign == NULL)
+ goto err_alloc;
- if (inet_add_protocol(&ipgre_protocol, IPPROTO_GRE) < 0) {
- printk(KERN_INFO "ipgre init: can't add protocol\n");
- return -EAGAIN;
- }
+ err = net_assign_generic(net, ipgre_net_id, ign);
+ if (err < 0)
+ goto err_assign;
- ipgre_fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel), "gre0",
+ ign->fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel), "gre0",
ipgre_tunnel_setup);
- if (!ipgre_fb_tunnel_dev) {
+ if (!ign->fb_tunnel_dev) {
err = -ENOMEM;
- goto err1;
+ goto err_alloc_dev;
}
- ipgre_fb_tunnel_dev->init = ipgre_fb_tunnel_init;
+ ign->fb_tunnel_dev->init = ipgre_fb_tunnel_init;
+ dev_net_set(ign->fb_tunnel_dev, net);
- if ((err = register_netdev(ipgre_fb_tunnel_dev)))
- goto err2;
-out:
+ if ((err = register_netdev(ign->fb_tunnel_dev)))
+ goto err_reg_dev;
+
+ return 0;
+
+err_reg_dev:
+ free_netdev(ign->fb_tunnel_dev);
+err_alloc_dev:
+ /* nothing */
+err_assign:
+ kfree(ign);
+err_alloc:
return err;
-err2:
- free_netdev(ipgre_fb_tunnel_dev);
-err1:
- inet_del_protocol(&ipgre_protocol, IPPROTO_GRE);
- goto out;
}
-static void __exit ipgre_destroy_tunnels(void)
+static void ipgre_exit_net(struct net *net)
{
- int prio;
+ struct ipgre_net *ign;
- for (prio = 0; prio < 4; prio++) {
- int h;
- for (h = 0; h < HASH_SIZE; h++) {
- struct ip_tunnel *t;
- while ((t = tunnels[prio][h]) != NULL)
- unregister_netdevice(t->dev);
- }
+ ign = net_generic(net, ipgre_net_id);
+ rtnl_lock();
+ ipgre_destroy_tunnels(ign);
+ rtnl_unlock();
+ kfree(ign);
+}
+
+static struct pernet_operations ipgre_net_ops = {
+ .init = ipgre_init_net,
+ .exit = ipgre_exit_net,
+};
+
+/*
+ * And now the modules code and kernel interface.
+ */
+
+static int __init ipgre_init(void)
+{
+ int err;
+
+ printk(KERN_INFO "GRE over IPv4 tunneling driver\n");
+
+ if (inet_add_protocol(&ipgre_protocol, IPPROTO_GRE) < 0) {
+ printk(KERN_INFO "ipgre init: can't add protocol\n");
+ return -EAGAIN;
}
+
+ err = register_pernet_gen_device(&ipgre_net_id, &ipgre_net_ops);
+ if (err < 0)
+ inet_del_protocol(&ipgre_protocol, IPPROTO_GRE);
+
+ return err;
}
static void __exit ipgre_fini(void)
if (inet_del_protocol(&ipgre_protocol, IPPROTO_GRE) < 0)
printk(KERN_INFO "ipgre close: can't remove protocol\n");
- rtnl_lock();
- ipgre_destroy_tunnels();
- rtnl_unlock();
+ unregister_pernet_gen_device(ipgre_net_id, &ipgre_net_ops);
}
module_init(ipgre_init);